To test the hypothesis that the apomorphine-induced increases in growth hormone are significantly negatively correlated with duration of illness (interval between age of onset and current age) and positively correlated with each other in both major depression and schizophrenia. Also, to correlate the apomorphine-induced decrease in prolactin with global ratings of severity of depression and specific depressive symptoms.